Search

Phospholamban (PLB)1 is a 52-amino acid transmembrane protein that interacts with the Ca-ATPase pump and lowers its affinity for Ca2+ (1-3). PLB plays a major role in the regulation process of the cardiac cycle (contraction and relaxation), which controls the heartbeat (3-5). Unphosphorylated PLB inhibits sarcoplasmic reticulum ATPase activity and stops the flow of Ca2+ ions, and this inhibition can be relieved by the cyclic AMP- and calmodulin-dependent phosphorylation of PLB (3-5). Since PLB is biologically significant and it is relatively small, many theoretical and biophysical experimental studies have aimed to investigate its structure in a membrane (6-22). On the basis of spectroscopic techniques and molecular modeling studies on pentameric WT-PLB, early structural reports on WT-PLB disagreed about whether the pentameric protein is composed of continuous α-helical subunits or composed of subunits that have two α-helices connected by an unstructured/β-sheet region (8, 21). Recently, the Chou group revealed an unusual bellflower-like assembly for pentameric WT-PLB that indicates an α-helical cytoplasmic domain of the pentamer that on average points away from the membrane surface (Figure 1A) (23). Alternatively, recent results from the Middleton group suggest that the cytoplasmic domain of PLB (residues 1-23) is stabilized only through its association with the phospholipid bilayer surface (24). In addition, new solution and solid-state NMR spectroscopic studies on the AFA-PLB mutant monomer (where C36, C41, and C46 have been mutated to A36, F41, and A46, respectively) have been reported (25, 26). The Veglia and Baldus groups debate whether the transmembrane domain of the monomeric AFA-PLB mutant structure is connected to an α-helical cytosolic segment that lies on and interacts with the membrane surface (L-shaped) (25) or a non-α-helical disordered cytosolic domain that has minimal interaction with the membrane surface (26). Figure 1 High-resolution solution NMR structures as well as the corresponding sequences for (A) pentameric (bellflower-like assembly) and (B) monomeric (L-shaped) phospholamban, by the Chou (23) and Veglia (26) groups, respectively. The two structures embedded ... Interestingly, the Veglia group resolved the solution NMR structure not only for the AFA-PLB monomer mutant (Figure 1B) but also for its phosphorylated form (27). The results of Veglia’s new phosphorylation study indicate that phosphorylation of serine 16 (located within the cytoplasmic domain) induces an order-to-disorder transition that disrupts the “L-shaped” monomer and causes a reduction in the extent of α-helical secondary structure around the phosphorylation site (27). However, the Chou group has not determined the phosphorylated form (P-PLB) of WT-PLB (23). They proposed that introducing a negative charge upon phosphorylation of WT-PLB could alter the average orientation of the cytoplasmic domain and decrease its accessibility to the calcium pump (23). Also, this latest model of WT-PLB interacting with SERCA introduced by the Chou group suggests that tightness of the PLB pentamer (supercoiled Leu/Ile zipper) could allow its cytoplasmic domain to fit into and interact with the groove of the SERCA cytoplasmic domain without the energetic cost of removing a subunit from the pentamer (23). Conversely, an earlier model suggests that phosphorylation of WT-PLB at Ser16 or Thr17 reduces its level of electrostatic binding with the pump and favors the association of the PLB monomers interacting with SERCA into PLB pentamers that do not interact with the pump (1, 3). Another recent model proposes that an allosteric interaction between the PLB monomer and SERCA takes place (28). Therefore, understanding the dynamic motion of WT-PLB with respect to the membrane is absolutely needed to completely describe the mechanism of binding of WT-PLB to Ca-ATPase. Complete knowledge of how different segments of phosphorylated and unphosphorylated PLB are moving within the membrane is essential for describing the mechanistic functions of the protein. EPR spectroscopic studies on the cytoplasmic domain of monomeric AFA-PLB have indicated large-scale dynamic changes (29). Also, Oxenoid and Chou hypothesized that this cytoplasmic domain motion is needed for WT-PLB/Ca-ATPase to function properly (23). It is clear that the cytoplasmic segment secondary structure, orientation, dynamics, and interaction with the membrane and SERCA of WT-PLB and P-PLB are under serious debate. However, it is important to emphasize that the high-resolution NMR solution structures of the monomeric PLB mutant (AFA-PLB) and pentameric WT-PLB in micelles reported by the Veglia (26) and Chou (23) groups, respectively, represent two of the best and widely accepted membrane protein structures available in the literature (see Figure 1). In addition, although the effect of phosphorylation on the cytoplasmic domain of WT-PLB has not been completely addressed in the literature, the Chou group suggested that their structure of WT-PLB in micelles sets the stage for a series of future experiments aimed at characterizing the effect of phosphorylation on the orientation and dynamics of its cytoplasmic helix (23). Our previous study indicated that phosphorylation does not significantly change the α-helical secondary structure of WT-PLB, and this finding agrees with the results previously reported by Arkin and co-workers (8, 30). In addition, that same report implies that P-PLB has less direct contact with the membranes than WT-PLB does (30). To extend and complement previous studies and to reveal new information about the dynamic properties of WT-PLB and its phosphorylated form (P-PLB), we have conducted NMR studies with PLB. These new studies will provide important physiological and mechanistic information regarding the regulatory role of WT-PLB and it phosphorylated form in biological systems (3, 23, 28). This dynamic information is crucial for understanding its binding to Ca-ATPase and inhibition. In this article, 2H and 15N solid-state NMR spectroscopic studies are utilized to ascertain pertinent information about the backbone and side chain dynamics of WT-PLB in phospholipid bilayers. 2H solid-state NMR spectroscopy is a powerful well-developed technique for studying the structural and dynamic properties of membrane proteins in phospholipid bilayers (31-34). The corresponding quadrupolar splitting and line shapes of the 2H solid-state NMR spectra can be used to probe the molecular dynamics of the side chain of selectively labeled residues in site-specific 2H-labeled integral membrane proteins (35-40). The primary amino acid sequence of full-length PLB has few alanines (mainly in the cytoplasmic domain, residues 1-15) and several leucines (the transmembrane domain, residues 23-52) along the length of the protein (23). In previous studies, methyl group motions have been well-characterized utilizing 2H NMR studies of CD3-labeled sites of alanines, valines, and leucines (35, 40-45). For the isotopically labeled alanines (short aliphatic side chains), the deuterated methyl group (CD3) rotates around the Cα-Cβ bond (see Figure 2A) and allows the deuterons to make jumps between three sites described by a tetrahedral geometry (46). However, for Leu, the long aliphatic side chain can be isotopically labeled at the δ- and/or ∊-CD3 sites, and the deuterium NMR powder pattern line shapes will be strongly influenced by the motions about the Cγ-Cδ bond axis as well as by additional librational motion about the Cα-Cβ and Cβ-Cγ bond axes (see Figure 2B) at various temperatures (40, 47). If the CD3-methyl probe of the protein undergoes no motion other than those associated with the axial rotation about the C-CD3 bond in a randomly dispersed sample, the resultant spectra will consist of a Pake pattern with a 40 kHz quadrupolar splitting (48). However, residues located outside the membrane are expected to be more motionally averaged and yield an isotropic peak (49). Figure 2 Chemical structure of the (A) alanine and (B) leucine amino acids. Additionally, 15N solid-state NMR spectroscopy has been used to ascertain direct information regarding backbone dynamics of membrane proteins (45, 50). Generally, the immobile (without large amplitude motions) amide sites of specific 15N-labeled proteins yield a broad static 15N powder pattern, whereas motionally averaged amide sites reveal isotropic peaks (50).

Cannabinoid receptors are G-protein-coupled receptors comprised of seven transmembrane helices. We hypothesized that the extended helix of the receptor interacts differently with POPC bilayers due to the differing distribution of charged amino acid residues. To test this, hCB1(T377-E416) and hCB2(K278-H316) peptides were studied with 31P and 2H solid-state NMR spectroscopy by incorporating them into 1-palmitoyl-2-oleoyl-sn-glycerophosphocholine bilayers. Lipid affinities of the 40- and 39-residue peptides were analyzed on the basis of 31P and 2H spectral line shapes, order parameters, and T1 relaxation measurements of the POPC bilayers. Lipid headgroup perturbations were noticed in the 31P NMR spectra in the lipid/peptide mixtures when compared with the pure lipids. 2H order parameters were calculated from the quadrupolar splitting of the de-Paked 2H NMR spectra. At the top of the acyl chain, pure lipids had an average S(CD) approximately = 0.20, whereas S(CD) approximately = 0.16 and S(CD) approximately = 0.18 were found in the presence of hCB1(T377-E416) and hCB2(K278-H316), respectively. S(CD) values decreased in the central part of the acyl chains when compared to the pure POPC lipids, indicating a change in the dynamic properties of the lipid membrane in the presence of the cannabinoid peptides. R(1Z) vs S2(CD) plots exhibited a linear dependency with and without the peptides, with an increase in slope upon addition of the peptides to the POPC, indicating that the dynamics of the lipid bilayer is dominated by fast axially symmetric motion. This study provides insights into the interaction of cannabinoid peptides with the membrane bilayer by investigating the headgroup and acyl chain dynamics.